Timestamps are in UTC.
sreynen, i think the key with the "you didn't really need to share that" type followups is to provide an alternative action to the impulse that the person had
i.e. rather than email such and such (...) to the list, it would help with efficiency, community memory etc. if you could do the following (...) on the wiki to capture the information in a form that is more easily found/referenced.
we're not going to be able to, nor do we necessarily want to, squash such impulses to contribute
however what we can do is direct them to more positive and efficient ends.
good idea, tantek
mailing list jujitsu
aikido even :)
[[plain-xml]] http://microformats.org/wiki?title=plain-xml&diff=0&oldid=13108 * Tantek * (+2652) added Doomed section and recent articles
[[xml-on-the-web-is-dead]] MN http://microformats.org/wiki/xml-on-the-web-is-dead * Tantek * (+30) add a redirect for a common phrase
[[xml-is-dead]] MN http://microformats.org/wiki/xml-is-dead * Tantek * (+36)
redmonk is Steve Ivy, http://redmonk.net and is linklogging at http://deliciouslymeta.com
pnhChris is Chris Casciano, blogs at http://placenamehere.com/ , and a member of the Web Standards Project.
[[depend-examples]] N http://microformats.org/wiki/depend-examples * DerrickPallas * (+4776)
[[depend-examples]] http://microformats.org/wiki?title=depend-examples&diff=0&oldid=13109 * DerrickPallas * (+23)
[[depend-examples]] http://microformats.org/wiki?title=depend-examples&diff=0&oldid=13110 * DerrickPallas * (+3) Portage: dev-lang/erlang -
tantek is Tantek <http://tantek.com> and works on Technorati and develops microformats <http://microformats.org>
[[events]] http://microformats.org/wiki?title=events&diff=0&oldid=13111 * Tantek * (-21) move past events to recent
Atamido is Paul Bryson, http://orangeman.commo.de/
Hey everyone, back from work. :)
mkaply is Michael Kaply <http://www.kaply.com/weblog/> and is the developer of Operator <https://addons.mozilla.org/firefox/4106/>
bergie is lives in Finland and blogs at http://bergie.iki.fi/blog/ and Midgard CMS developer
Hi there bergie.
Any exciting events transpire today?
Atamido is Paul Bryson, http://orangeman.commo.de/
bergie is lives in Finland and blogs at http://bergie.iki.fi/blog/ and Midgard CMS developer
sreynen is Scott Reynen, who makes things at makedatamakesense.com
Charl is Charl van Niekerk and writes about standards at http://standards.za.net/
bengee is Benjamin Nowack (http://bnode.org/)
trovster is a web developer from the UK who writes on http://www.trovster.com and helps with www.multipack.co.uk
iand is Ian Davis who blogs at http://iandavis.com/blog
[[species]] http://microformats.org/wiki?title=species&diff=0&oldid=13112 * AndyMabbett * (+592) Proposal - another problem/ use-case
[[species]] M http://microformats.org/wiki?title=species&diff=0&oldid=13113 * AndyMabbett * (+2) ).
[[species]] M http://microformats.org/wiki?title=species&diff=0&oldid=13114 * AndyMabbett * (+0) Proposal - typo
[[species-examples]] M http://microformats.org/wiki?title=species-examples&diff=0&oldid=13115 * AndyMabbett * (+3) Plant - fmt
[[profile-examples-in-wild]] N http://microformats.org/wiki/profile-examples-in-wild * AndyMabbett * (+416)
[[hcard-profile]] http://microformats.org/wiki?title=hcard-profile&diff=0&oldid=13116 * AndyMabbett * (+29) Related Pages - profile-examples-in-wild
[[hcard-profile]] http://microformats.org/wiki?title=hcard-profile&diff=0&oldid=13117 * AndyMabbett * (+31) Related Pages - profile-examples-in-wild
[[profile-uris]] http://microformats.org/wiki?title=profile-uris&diff=0&oldid=13118 * AndyMabbett * (+61) Profile examples, in the wild
veeliam is William Lawrence <http://zaxbypass.com>
[[this-week-pending]] http://microformats.org/wiki?title=this-week-pending&diff=0&oldid=13119 * AndyMabbett * (+129) Profile examples, in the wild
[[this-week-pending]] M http://microformats.org/wiki?title=this-week-pending&diff=0&oldid=13120 * AndyMabbett * (+1) This Week in Microformats - Pending Items - .
[[Main Page]] http://microformats.org/wiki?title=Main_Page&diff=0&oldid=13121 * AndyMabbett * (+61) Tools, Test Cases and Additional Research - *Profile examples, in the wild
[[events]] http://microformats.org/wiki?title=events&diff=0&oldid=13122 * AndyMabbett * (+29) Recent - correct date order
[[events]] http://microformats.org/wiki?title=events&diff=0&oldid=13123 * AndyMabbett * (-9) Fix alignment of edit links ("H1" not needed if NOTOC used)
[[events]] M http://microformats.org/wiki?title=events&diff=0&oldid=13124 * AndyMabbett * (-20) Microformats related events - meanigngful link text
[[events]] M http://microformats.org/wiki?title=events&diff=0&oldid=13125 * AndyMabbett * (-1) Microformats related events - tyop
Charl is Charl van Niekerk and writes about standards at http://standards.za.net/
http://www.brucelawson.co.uk/index.php/2007/silly-microformats-question/ anyone got more suggestions for this.
there was an openign hours suggestion on the wiki that said use vevent
http://microformats.org/wiki/to-do#for_all_microformats
julianstahnke is Julian Stahnke and works for last.fm and implements microformats wherever he can
Ronnos is Ron Kok, a friendly student Communication and Multimedia Design in The Netherlands
Kilianvalkhof is Kilian Valkhof, he makes websites and blogs at http://kilianvalkhof.com
mflogbot is logging #microformats (http://microformats.org/wiki/mflogbot)
Mr_Elusive is not a programmer from id but makes his home at http://eswat.ca
ping
pong :)
there are some people alive this time of day - i'm from south africa
pong the pint
ping
I jsut wanted to test if mflogbot is logging again ;-)
mkaply is Michael Kaply <http://www.kaply.com/weblog/> and is the developer of Operator <https://addons.mozilla.org/firefox/4106/>
edsu is Ed Summers from the Library of Congress <http://www.inkdroid.org>
RobertBachmann is Robert Bachmann <http://rbach.priv.at> from Austria (TZ: 0100)
can anyone think of a logical reason why running xslt transforms locally would be different than those same transforms run on a web page?
different xslt engine?
which xslt?
whafro is M. Jackson Wilkinson, a designer/developer for Grassroots Enterprise in Washington, DC
RobertBachmann: I downloaded Brian's X2V and am integrating it into Operator for debug purposes
So you can compare his results to mine.
It's working in some cases, but for instance it's not handling the title tag in microformats
really weird stuff
which xslt engine do you use?
Expat. built into Firefox. I didn't think about that. He's using a server based engine
so it might be a bug in the Firefox XSLT implementation. duh
let me check tails export
no exapt is an XML parser.
Same problem with tails (which uses X2V).
OK, going to check what xsLT parser they use
Transformiix
Gee. I guess I have a better answer as to why I didn't use Brian's work for Operator :)
[[species]] http://microformats.org/wiki?title=species&diff=0&oldid=13126 * AndyMabbett * (+82) Proposal - examples of common words in taxons
[[species]] http://microformats.org/wiki?title=species&diff=0&oldid=13127 * AndyMabbett * (+34) Proposal - another example, & fmt
csarven is Sarven Capadisli and can be found online at http://www.csarven.ca
whafro is M. Jackson Wilkinson, a designer/developer for Grassroots Enterprise in Washington, DC
redmonk is Steve Ivy, http://redmonk.net and is linklogging at http://deliciouslymeta.com
SamRose is found at http://smartmobs.com, http://communitywiki.org, http://blog.p2pfoundation.com, http://barcampbank.com, and http://cooperationcommons.com
http://bergie.iki.fi/blog/contact_management_and_microformats.html
pnhChris is Chris Casciano, blogs at http://placenamehere.com/ , and a member of the Web Standards Project.
mkaply is Michael Kaply <http://www.kaply.com/weblog/> and is the developer of Operator <https://addons.mozilla.org/firefox/4106/>
mlinksva is Mike Linksvayer and from Creative Commons
[[recipe-brainstorming]] http://microformats.org/wiki?title=recipe-brainstorming&diff=0&oldid=13128 * JohnLeMasney * (+153) Suggested fields for inclusion -
[[recipe-brainstorming]] http://microformats.org/wiki?title=recipe-brainstorming&diff=0&oldid=13129 * JohnLeMasney * (+183) Scope -
danja is Danny Ayers, http://dannyayers.com
any of you guys have a blogger account? i need to test atom feed autodiscovery
Hi everyone.
briansuda is brian suda of http://suda.co.uk and is at (-0000 GMT) and is author of "Using Microformats" for O'Reilly [http://www.oreilly.com/catalog/microformats/]
bergie is lives in Finland and blogs at http://bergie.iki.fi/blog/ and Midgard CMS developer
mmc_ is Michael McCracken and can be found online at http://michael-mccracken.net/
hm, there is a hReview format, but where is the hPreview?
[[geo-cheatsheet]] M http://microformats.org/wiki?title=geo-cheatsheet&diff=0&oldid=13130 * AndrewKuchling * (+1) Notes -
redmonk: still there?
It's all about hPostview
We need a format that allows us to tell us what we have just looked at.
mkaply, i emailed you RE: Firefox ABBR@title issues
briansuda: Just saw. Tx. I'm talking with one of the Mozilla XSLT folks now
let me know what transpires
briansuda: That explains why I was seeing such weird results when I compared my stuff against yours :)
hey briansuda, do you know if there is something like an hPreview?
to Preview something? what are you trying to accomplish?
the rating in hReview is optional, so you can use that and just not rate it
well, on my website i started to review books, but there are new books comming up, and some of them i want to preview
well, what semantic data would you want to capture about these previewed books?
well, release date for sure
that can be an hCal event?
i know it's almost the same as a hReview
[[recipe-brainstorming]] http://microformats.org/wiki?title=recipe-brainstorming&diff=0&oldid=13131 * AndyMabbett * (+126) Scope - calorific values
the place where you previewed it? as an hCard?
hm...
well, the only difference would be the release date to indicate it's not in stores yet
then you are reviewing an event in the future
something like that :S
hReview can have a dtpublished of today and talk about an event in the future
so the only problem would be the format name :)
why?
because it's not possible to review something that did not happen yet?
no, you can review an upcoming event.
well, you're reviewing the preview copy of the book/game/record
that is exactly what a pre-VIEW is doing
you are not rating it... that would be impossible to rate it without it happening
is type in hreview a list of types or can you add your own?
[[species]] http://microformats.org/wiki?title=species&diff=0&oldid=13132 * AndyMabbett * (+230) Proposal - more
the wikipage shows the following options: product | business | event | person | place | website | url.
but i think its extensible?
sounds pretty simple: hreview with release date in the future
it is an enumerated list
Ronnos - see hreview-faq re types
briansuda: Nothing jumps out at the XSLT guy. How hard would it be to do a reduced testcase ? :)
thnx tantek
mkaply, we can do a reduced test case, what do you have inmind
briansuda: Just something simple that runs against an abbr and tries to get the title. To see if that is the problem.
Using the same method you are using. Everything else seems to work locally. The only problem I see is that no titles are being extracted from abbrs
we might have a test case for that already in hg - let me check
how about this one: http://hg.microformats.org/tests?f=2bf906aaa71d;file=hcard/25-geo-abbr.html
bewest, my problem is, if i ad hCalendar to the hReview, it looks like i'm reviewing (or previewing, depending on the date) the event, in stead of the item releasing on that date
I was more referring to the reduced XSLT to try to see what transformiix is doing wrong.
ah, hm.. ok
do you want me to email you something?
briansuda: If you could that would be awesome. I know nothing of XSLT. Basically just looking for a "here's an XSLT that when run against this ABBR shows the problem"
Ronnos: why?
ok, give me a few minutes.
I don't see why putting a release date in a feature all of a sudden makes it appear that the review is of an event instead of the item being reviewed
oops, "in the future" not "in a feature"
maybe i messing thing up, need some coffee :)
i don't know why i did, but i was thinking that i only could place an hCalendar event in "item"
ah, I see. because there is not datereleased field
I wonder what hproduct is doint
doing
it's describing a product
who made it, how it looks, where you can find it on the web
yes, but the fact that you're previewing an item carries the same semantics for hreview. the only difference is that the release date is in the future
release dates are a property of a product, not a review
thats true
so you're best bet is to continue using hReview
and help out the hProduct guys with their work ;-)
and use the include pattern in the "item"
:P
because it looks like hProduct has a ways to go
*in/for
and I'm sure extra help would be appriciated
especially since you have examples ready to be published
you could be giving them great feedback
love to help, the problem is that i think in dutch :P
lols
Ronnos: the guys involved with that are Aaron Gustafson, Adam Craven, and Craig Cook
Ronnos: in fact, you're input would be very helpful, because my brief glimpse of their work does not include the release date
Ronnos: so you would need to present a clear case why including release date is a good idea
ow wooow... you indicate the date a product was released with hCalendar, you include that date in hProduct, the product was made by someone, that would be an hCard contact, you can review it with hReview
*yoinks*
dunno if hcard is necessary to link the hreview to the product
there will probably be a way to directly link an hreview and hproduct without needing to go through who made it
that's what I would expect anyway
but currently, there is no mechanism to specify when a product was (or will be released)
it would clearly use hcalendar, but the mechanism that specifies that the hcalendar signifies the release date hasn't yet been suggested
so if you want that, I suggest you bring it up on the mailing list or their wiki
bewest, what is the best way to start contributing?
be sure to include your own use case for doing so
and it would also help if you could find some websites where people publish reviews for future products
maybe engadget or something?
Ronnos: the best way to contribute is to share your need to publish something
amazon.co.uk does it
with the people creating the format
mailing list, wiki, are good forums, as is the IRC chat room and a blog
if it's not obvious that your need would contribute to the overall use case, you may be asked to gather examples of other people publishing in the same way you intend to
eg, amazon, like you just mentioned
can you find a specific example of this on amazon?
any specific example from as many sites as you can would be extremely helpful
Morning all.
morning
mkaply, emails away
briansuda: you rock
i just hope it is useful
briansuda: Of course you couldn't test if it failed in Firefox transformiix, could you :)
walk me through it and i can (maybe)
briansuda: I might have to write a quick extension for it. Give me a few min
well, amazon.co.uk displays the books for preordering, and in some cases the discribe the book and it's contents. It looks like they've disabled the review option
I think engadget does reviews of products that aren't released to market yet
[[product-brainstorming]] http://microformats.org/wiki?title=product-brainstorming&diff=0&oldid=13133 * Ronnos * (+420) Random Thoughts -
bewest, how about gaming sites!
they preview (and even rate) all the time
Ronnos: it helps to list specific URLs :-) gaming sites are a great idea
good idea
"The message's content type was not explicitly allowed" WTF?
eh, bewest, where can i list those examples?
the examples page
product-examples
if you collect some URLs, I'll help you out a bit by telling you which bits are the good parts to explain
bewest, have a look at this:
http://www.amazon.co.uk/s/ref=nb_ss_w_h_/202-2502647-9229460?url=search-alias%3Daps&field-keywords=microformats&Go.x=0&Go.y=0&Go=Go
it's a good one ;) about Microformats :)
it's a product page, and the release date is in the future
indicating it's not available yet
[[User:Ronnos]] http://microformats.org/wiki?title=User:Ronnos&diff=0&oldid=13134 * Ronnos * (+6)
Ronnos: excellent
i wonder why the hProduct format isn't fineshed yet, it seems to me like an important one?
"finished"
:)
Ronnos: it's kind of like cooking
Ronnos: what happens if you rush food?
;)
so one thing to take notes on is simply the list of properties being published
we don't necessarily need to look at the source for that
it's pretty obvious in this case
true
name of book, name of author, release date, availability status (unavailable), availability options (used vs new), list price, selling price, medium (paperback)....
can you see any others?
picture/image?
you know, for existing formats should looj at google checkout
bewest: http://videogames.netscape.com/story/2006/10/03/halo-3-available-for-pre-order-on-amazoncom
yes, picture/image good catch
hold on
not done with amazon yet
time to look at the source
interesting
<h1 class="breadCrumb"> "microformats" </h1>
ah, that's just the search
yes, there are some more properties published which aren't apparent unless you look at the source
they are also publishing category information
this product is in the category of "books"
yup
so in their publishing system, category is a property of the item
class="imageColumn"
some semantic markup for declaring the image
<span class="srTitle">Microformats: Empowering Your Markup for Web 2.0</span>
<span class="bindingBlock">(<span class="binding">Paperback</span> - 26 Mar 2007)</span>
ok lunch is here
http://www.atomicbooks.com/results.php?ep=49&cat=49
Ronnos: this is the kind of information to be collected on the -examples
page
the provide a date for the scheduled arrivel
arrival
neat
ah okey
thnx
thanks
:-)
add as much as you can to the examples page
:-)
okay, gonna try to make myself usefull :)
i'd like to get the membership of the -admin list documented in the wiki so we can stop seeing hyperbole like "secret police"
This FAQ seems like the best place to put a list: http://microformats.org/wiki/faq#Q:_Who_controls_microformats.3F
does anyone mind if i start a list there, without including anyone who hasn't already publicly declared their participation on the -admin list?
tantek, kingryan, bewest, briansuda, KevinMarks?
you can add me to the list, i'm fine withthat
[[product-examples]] http://microformats.org/wiki?title=product-examples&diff=0&oldid=13135 * Ronnos * (+227)
What's everyone up to this afternoon/evening/morning?
[[faq]] http://microformats.org/wiki?title=faq&diff=0&oldid=13136 * ScottReynen * (+249) Q: ''Who controls microformats?'' -
okay, i listed the people who have already declared their membership in public. please edit the wording if you see any problem with it
sreynen: yeah do it
oh too late
ummm
good job
heh, thanks
what's interesting to me is that some people appear to think that evidence that shows that a technique is working is also evidence to change the technique
joe's issue seems to come down to andy's messages being delayed
i think the list is used too much for real-time back and forth anyway, but that's a separate issue entirely
hmm
sreynen: did you catch my suggestion on the admin list to frame the group's efforts around capturing techniques and then only when needed creating new formats?
it seems what most people are requesting on the list is how to do something, for which the building blocks are already present
however, they understand this to be, primarily, a formats group
and so they ask for a new format
instead of asking how to accomplish a certain feat
yeah, i think that's a good point
^^
the entire process is centered around creating new formats
which is why so many people ask for them
bewest, adding yourself to the Autors part on an example page if you added some examples... is that allowed? Or does it serve another purpose
Ronnos: yes, please add your self
bewest, thnx
Ronnos: did you see earlier how I was starting to look at the actual markup?
Ronnos: do you think you could add some similar markup analysis to the wiki page?
yeah, i supply class information
i name the sort of information they give, and also include the class name they use to markup the information
ah, nice
I was especially curious, becuase there is a claim that amazon is publishign this stuff in an "unstructured" fashion
whereas when I took a look at it, there was lots of semantic encoding
this is a good sign, even if hProduct class names end up being different, because it shows a receptivity and aptitude on the part of publishers, and confirms the need to "pave that cowpath"
right
a lot of sites are including a big list of product features
and a lot of features are depending on the kind of product
hmm, yes
features, type, value
I've had some ideas on that, but I wanted to run my idea past DanC first
should probably just try it and see what happens
make a kind of dummy
almost every product has a Specification list
be careful :-)
don't "fall down the rabbit hole" trying to come up with a general abstract mechanism of specifying structured data
the reason we list what each publisher publishes is so we can find out which properties are the most common
uhuh
we'll be satisfied when we can produce a list that conforms to 80/20 guideline
[[product-examples]] http://microformats.org/wiki?title=product-examples&diff=0&oldid=13137 * Ronnos * (+322)
[[product-examples]] M http://microformats.org/wiki?title=product-examples&diff=0&oldid=13138 * Ronnos * (+15)
briansuda: are you ready?
ok, for a few minutes, then i have to find some food
:)
briansuda: It's a case sensitivity issue with your transform.
In operator I clone the node and firefox is uppercasing all the ABBR tags
abbr tags
abbr vs. ABBR
yep
ok, XML is case-sensitive
and all the elements are lower-case
so this sounds like an issue with FF
(all the elements in XHTML)
briansuda: except that microformats can be in HTML, so should your transform be case sensitive?
he strange part is everything else about your transform is working.
hm. well i run everything through TIDY to make it XHTML
so it will be converted from <P> to <p>
but this still doesn't make sense in some way
what specifically?
the original node was correct. I clone the node. I never serialize it
so is Firefox cloneNode really modifying the nodes to uppercase the nodeName? That would be crazy
yep. That's what it is doing
i think you found your bug
cool.
:)
feel free to email me if you find anything else, and/or need more test cases
i'm going to find some food
briansuda: I really appreciate your help
not a problem, that's what i'm here for
it would be silly for me to sit around IRC and not do anything
tantek: the moderation complaint...
http://wiki.mozilla.org/Firefox:3.0_DOM_Case_Sensitivity
Ronnos, I just read that thread and it had only 11 items. There's got to be more going on to cause the 100+ emails.
in a matter of days
tantek, is the google group in sync with the mailinglist?
the google group is not run by the list admins
don't trust it
what google group?
http://groups.google.com/group/microformats/
tantek: several format proposals, along with several questions about hcard, along with several questions about parsing rel-tag
how odd
the google group
messina set it up
weird
or at least redundant
I remember him doing it
well, to me it looks up to date?
a fork in progress
an accidental fork
thats easy, just call the boys at google ^^
I'm sure it's easier than that
:)
hey mkaply: is it possible to automate the tests you run against operator?
bedtime, i call it a day
Hey folks!
What's shake'in?
kingryan: most of the tests I run are "go to this page" "see if the action works"
But I'm working on the ability to pass a microformat to the parser and see if it validates
not automatable?
in the sense that you can look at the javascript output and see if the values are what you expect
I've never done firefox UI automation specifically.
What exactly are you looking to see from a testing perspective?
well, I'm looking to start automating testing of the major mf parsing implemenations
so that we can compare them side-by-side
kingryan: taking cues from the OWL efforts?
yeah
I'm motivated 'cause i'm looking at x2v right now and realize that it's failing a number of tests
if someone pushes something that doesn't pass, mf-dev should get an email
mkaply: read Dan Connolly's blog for an analysis on how OWL did testing
mkaply: http://dig.csail.mit.edu/breadcrumbs/node/171
then take a look at the OWL testing pages
I think using the parser I'm working on it could be automated. Essentially you would have a page that had the expected values. You would invoke the parser under the covers on a given microformat. It would return the data.
Then you could compare the expected output vs. what we report in the JS structure.
sounds like a good start, yeah
Essentially all Operator is doing is parsing the microformat into a Javascript representation of the mf
I'm actually in the process of working with Andy Mitchell to try to create a standardized JS parser that anyone could use
that's good...
what if you want to compare your results to an implementation that isn't in javascript?
JSON!
:-)
http://dichotomize.com/czmap/hcard_structure.js
well you could take the JS output and transform it to whatever you want.
kingryan and crew: Congrats on WTF.
mkaply: assuming certain things about the js output, correct?
thanks TylerR
bewest: yeah. We have a definition file that in theory shows exactly how the JS output will be
properties/subproperties - which will be arrays/objects/strings
mkaply: I thought your definition mixed code and data
bewest: it does.
isn't that a problem for comparing output with non JS implementations?
bewest: well, you can compare the data nodes only
mkaply: so it sounds like it might be worth decoupling the implementation from the data representation
bewest: we thought about that at first, but that's not useful for validation
your template doesn't say which fields are ISO dates for instance
what does enumerable mean in your context?
knowing what a representation means is irrelevant in terms of comparing like representations
that's something for the parser
enumerable...
I forget
let me see if I can mind read myself
the structure you present comes very close to what our parser outputs.
enumerable means that the possible values are enumerated
mkaply: yeah, form somewhat correlates to function :-)
ok. I saw lots of enumerable false except on additional-names and and the honorific stuff
mkaply: but my concern is mixing code and data
how do you specify that a property can only be singular or multiple?
bewest: I totally understand.
what do you mean?
I think you don't need to specify singular or multiple
because rules for parsing indicate it
See when I'm thinking template, I'm thinking someone can give me a template and I will know how the details of how to parse the microformat.
Knowing what is a child of what isn't terribly useful.
My thinking (and I could be wrong)
was that we could create a scenario where someone could tag up a definition file with very little javascript (the getters only handle exceptions and specialized types)
and with that definition file have full support for that microformat.
so there are two things here. There is a JSON template which can be used for simple validation
and there is what we are calling a definition file which handles all the details of taking any given microformat and turning it into our internal JS structure
See in my mind, you can'
a property could be added to the meta field to describe ocurrence rules
See in my mind, you can't represent nickname as one field in a structure. because it could be an array.
and if you represent it as one field, then you need a separator, and then things go wonky :)
maybe ocurs: "singleton" | "multiple"
an array?
it evaluates to a single value
no, I can have six different nicknames. IT shouldn't evaluate to six nicknames separated by a space
the evaluation itself is procedural
are people publishing six nicknames?
well no :)
I just use semi-colons.
[[hcalendar-implementations]] http://microformats.org/wiki?title=hcalendar-implementations&diff=0&oldid=13139 * Glasser * (+26)
Who has a semi-colon in their nickname?
but when I'm communicating with web services, I might only be able to pass the first instance of something.
If you do, I don't want to be your friend.
So I need to be more specific.
then pass $list.split(";",2)
but what about my friend B;ob ? :)
whose name is B;ob?
I think he's got bigger problems
Yeah, everyone mistakes him for B:ob.
there are people that use colons in their nicknames though. that's funny
Poor old "ob", the B namespace never treated him right.
but seriously. There is a way to separate things.
What I can do is have a separate getter array.
then the main structure would have no code.
so, is there a unit test with more than one nickname?
It isn't a singleton.
It's also very easy to deal with.
most things aren't singletons
It's an implementation detail.
nicname is singleton
in an hcard
Is it?
Oh, then just pull the first one.
http://microformats.org/wiki/hcard#Singular_vs._Plural_Properties
Simple.
For properties which are singular (e.g. "N" and "FN"), the first descendant element with that class should take effect, any others being ignored. For properties which can be plural (e.g. "TEL"), each class instance should create a instance of that property.
There is a test for people with multiple n.
not N, NICKNAME
right, that's what I was saying the evaluation is procedural
but nickname is a property of N
actually it's not
it stands alone
hmm
oh wait
Note: the hCard may have additional explicit "nickname" property values in addition to the implied nickname.
mkaply is right
Which of the sub-properties of n are singluar?
so is there an hcard test with multiple nicknames?
there is no nickname test.
anyway, mkaply, you raise an interesting issue
however,
Our goal is to create something that is useful to the microformat community, so I want as much communication as humanly possible.
if you are communicating with some interoperable web service, it would be advisable to pass messages in a representation that is intended to be interoperable, rather than a representation which is the result of a transformation from one that was interoperable
Singular properties: "FN", "N", "BDAY", "TZ", "GEO", "SORT-STRING", "UID", "CLASS".
Nickname is not a singular property.
right
And nickname is not a property of n.
bewest: But my point is I would much rather have an internal representation of a plural property that keeps the pluraliry. So I can determine the separator based on the service I am giving the data to
someone want to write a nickname test?
Which is why it confuses me that you imply it iff n is not there.
DerrickPallas: basically they are saying if fn is one word, it's probably a nickname.
Does that mean that n:=nickname if !n and nickname?
Which personally I think is stupid
http://diveintomark.org/projects/greasemonkey/hcard/tests/
No, I understand what they're saying.
http://diveintomark.org/projects/greasemonkey/hcard/tests/3-1-3-nickname-2-plural.xhtml
I think the wiki has a typo..
DerrickPallas: where?
Under Implied "nickname" Optimization
Similar to the implied "n" optimization, if "FN" and "ORG" are not the same, and the value of the "FN" property is exactly one word, and there is no explicit "N" property, then:
should be
Especially since I haev an Indian friend whose name is one word - Ravisankar
Similar to the implied "n" optimization, if "FN" and "ORG" are not the same, and the value of the "FN" property is exactly one word, and there is no explicit "NICKNAME" property, then:
That's not his nickname
kingryan: do you think that's an error on the wiki?
http://dev.w3.org/cvsweb/2001/palmagent/hcardTest.html?rev=1.13&content-type=text/x-cvsweb-markup
same construct
<ul class="nickname"> <li>Jim</li> <li>Jimmy</li> </ul>
DerrickPallas: error, no. bad idea, maybe
that's invaid
the nickname would be JimJimmy
hmmm
How do I view that HTML as HTML?
That's not the question.
The wiki says:
the wiki doesn't make any sense to me on this subject
it looks like a copy-paste gone wrong
imply that nickname:=fn under the following circumstances
bewest, DerrickPallas : url?
* fn != org
* fn contains a single word
* n is not given
http://microformats.org/wiki/hcard#Implied_.22nickname.22_Optimization
yes. that is exactly what it says. I actually understand the logic.
DerrickPallas: I think you're probably right
I'll edit it then.
well, that's one possibility
wait, doesn't it say that nickname is added to the list of nicknames?
there is another possibility as well, correct?
mkaply: that's a different question
wait, no
that implies that nickname has been set, so that last part can't be incorrect
"# Parsers should handle the missing "N" property by implying empty values for all the "N" sub-properties. "
mkaply: it says that there can be additional nicknames, we're questioning a different part
it just says N should be empty
The reason they are bring N into the equation is because they are saying that it's not a "real name" becuase there is no n
That's always been my assumption
we need to ask tantek about this
ok
he's the one who wrote it
http://corkd.com/wine/view/8001
I believe this is an example scenario in this case. Take a look at these contacts
what's really weird about it is that nickname is not a sub-property of n
DerrickPallas: but n is important in this situation. I believe they are saying is that with no N, it must be a nickname or they would have put info in N
mkaply: we're not discussing implementation, we're discussing why the standard is written that way
so the evaluation of nickname is: 1.) the implicitly optimized value consisting of the value of n-optimization when fn and org are not the same and n is absent and fn is one word - AND - 2.) the additional explicitly listed nicknames
clearly that wasn;t the intent
or this line wouldn't be present
Note: the hCard may have additional explicit "nickname" property values in addition to the implied nickname.
that's what the spec indicates
that is # 2 in my list
the spec says implied nickname is always a part of evaluating nickname
sorry, I read it wrong. I thought you were going back to the "they meant nickname not N"
which is why I listed it as number 1
oops there's an otherwise on #1
if fn is one world, you use the n-optimization
no
if fn is one word you use fn
This so shouldn't be in the spec. Just because an fn is one word doesn't mean it is nickname
That's very eurocentric
if fn is not one word, and fn isn't the same as org, you do the n-optimization and use that value for nickname
mkaply: should or shouldn't be is a different question than "what does this mean"
to determine the apropriateness, we have to determine what it means first
mkaply: yeah, avoid passing judgement, we're just trying to decipher it
bwest: I thought the nickname optimization only ever happens if fn is exactly one word.
nickname optimization always happens
the procedure differs depending on whether or not fn is one word along with several other conditions
if if "FN" and "ORG" are not the same, and the value of the "FN" property is exactly one word, and there is no explicit "N" property, then
Those are the conditions to cause the nickname optimization
no
those are the conditions to evaluate nickname optimization
Ryan, can you ask Tantek about the conditionals?
mkaply: "Note: the hCard may have additional explicit "nickname" property values in addition to the implied nickname." means nickname always has an implied value
DerrickPallas: tantek's around in IRC
so if it said "to an implied nickname" it would be OK?
so just to restate my interpretation: values for nickname = impliedNickname + explicitlyListedNicknames
impliedNickname = if(fn != org && noExplicitValue(n)) { if(isOneWord(fn) { return fn; } else { return optimizeN(); }
Where does it state that the implied nickname is N?
I'm sorry, I hate to be dense but I don't understand where you are getting the else { return optimizeN(); } from the spec
It's not there.
ah
impliedNickname = if(fn != org && noExplicitValue(n)) { if(isOneWord(fn) { return fn; } else { /* no return */ optimizeN(); }
so... it doesn't make any sense to me
hehehe
Statement one says to pretend like nickname is fn. Statement two says nothing about nickname, only about N.
It is identical to the statement in "Organization Contact Info" which alreadys says that empty N means all the sub properties are empty.
yes, that's what my pseudocode does
From now on, I'm going to write "optimizeBen" in all my code and make a note someplace in another file that optimizeBen should evaluate to the empty string if it evaluates to the empty string.
And it always will.
:)
At least whenever it gets used.
actually, it seems to me that it just says when no nickname is provided, when fn and org are not the same, when N is absent, let nickname equal N optimization, but without org. Also, let N = n optimization
I agree it's a silly way to say things
It doesn't say anything about no nickname being provided.
fine
It says the opposite, that other nicknames might be provided.
yes yes
nickname is all the nicknames provided plus an implied nickname described by "when fn and org are not the same, when N is absent, let nickname equal N optimization, but without org. Also, let N = n optimization. Also since n is absent, it's empty."
lol
Instead of just saying the word empty, I'm going to say "flibbertygibbet" from now on; where currently, flibbertygibbert is always the empty string, with the caveat that it may change in the future.
Just to make things more confusing.
well, which is more confusing? the way I stated it or the way it's stated in the spec
either way, it doesn't make any sense
especially if they mean the same thing
yes the "N is empty" condition re: nickname is confusing, and probably was unintended
They superficially and coincidentally mean the same thing.
that's all I was going for :-)
Oh, golly, how could I stay mad at you?
DerrickPallas also just pointed out there's no conformance guidelines for parsing of adr
If I see an adr with no sub-properties, I'm just going to use the flattened value of the element.
yeah, I guess I'm leaning that way too
If it has sub-properties, do I only use the flattend values of the sub-property elements?
These logs were automatically created by mflogbot on chat.freenode.net using a modified version of the Java IRC LogBot.
See http://microformats.org/wiki/mflogbot for more information.